How to Do a Good PowerPoint Lecture zz
Having strongly stated my opinion that PowerPoint is not actively evil, but can be used to give good scientific presentations as well as soul-crushingly dull bullet-point talks, I feel like I ought to say something to back it up. Here, then, are some of the rules of thumb I use when putting together a good PowerPoint talk.
1) Know Your Audience. This is probably the most important rule in giving a talk, no matter what medium you plan to use. A talk aimed at an audience of undergraduate science majors is a very different thing than a research talk given at an international conference, or a public lecture given to a general audience. Don't assume you can splice one of these together out of pieces of another-- in certain circumstances, you can re-use slides from a talk aimed at one audience for a different kind of audience, but it's not a safe assumption. You're better off starting from scratch.
2) Limit Your Material. I tend to view one slide per minute as an absolute upper bound on any given talk, and I rarely reach that. The most recent public lecture I've given ran twenty slides for a 50-minute talk, and I rarely have more than 15 slides for a one-hour class. The colloquium talk on my current research runs about 26 slides, and also tends to run a little long.
I've seen more speakers than I care to mention show up with 75 slides for a 60-minute talk, and those talks never end well.
3) Equations Are Death. Yes, physics is a mathematical science. It doesn't mean that people want to look at slide after slide of nothing but equations. There are few experiences as soul-crushing as sitting in a dark room watching somebody do algebra.
It's almost impossible to completely avoid equations, but make sure you limit it to only the absolute essentials, and do everything possible to simplify them. If your equation has a big clump of fundamental constants out front, consider grouping them together into a single overall constant, to redue the visual clutter.
4) Text Is Death. Specifically, don't put complete sentences on your slides. Most people can read faster than you can talk, and if you present them with a dense slide of text, they'll focus on reading it, and tune you out. You want them to be listening to you, not reading your slides, so keep the text to a minimum. You need enough that the slide will make sense to somebody who gets distracted for a minute, and misses some of the talk, but you're not writing a novel, you're giving a talk.
5) Explain Your Graphics. When you show a plot, explain what it is clearly and simply, the minute you put it on screen. Say right up front "In this graph, we show the number of neutrinos at a given energy on the vertical axis, versus the neutrino energy on the horizontal axis." If the audience doesn't know what you're plotting, they won't understand anything else that you say, so make it clear from the beginning.
6) Define Your Terms. Similarly, if you must put up equations, make sure you define all the symbols clearly and simply. Even if something is just a proportionality constant, say that up front, otherwise somebody in the back is going to be thinking "What the hell is A?" and missing your actual point.
7) Keep the Background Simple. Yes, you can download all sorts of spiffy pictures to use as a background for your slides. Don't. It's visually distracting, and tends to obscure your actual meaning.
The background should be a solid color, or if that's too plain, one solid color shading into another. The text and graphics should be in colors that strongly contrast with the background-- yellow text on a white background is almost completely invisible, and should never be used. You want people to be listening to you, not squinting at the screen trying to determine whether there are words there.
8) Keep the Animation Simple. There's some setting in PowerPoint that assigns a random animation effect to each new item that appears, and students almost invariably go for this. Whenever I see it, I want to beat them senseless with an eraser.
You can get a lot of mileage out of simple animation, revealing important information a little at a time, but the carnival of animation effects is incredibly distracting. Pick one effect, and stick with it. I recommend "Appear," but I don't object to some of the fade-in effects. The one where a text item drops from the top of the screen and bounces three or four times before settling into place should never, ever be used except as an "alternative interrogation technique."
9) Keep Multimedia to a Minimum. One of the selling points of PowerPoint is that it lets you stick in movies and sound files, and other multimedia objects. Resist the temptation to overuse this feature.
The sad fact is, these files will fail to work about a third of the time. Two-thirds of the time, if you're using somebody else's computer. Don't set up your talk so that it relies on the seamless operation of some video clip that requires a plug-in, because it's almost guaranteed to fail at some point. If you've got to show a movie, make it an animated GIF if you can, rather than trusting that QuickTime will work when you need it. And put in a slide showing still images of the key frames, in case the movie just won't play.
10) Prompt Yourself. There should never, ever be a time when a slide comes up and you think "What the hell is this?" Make sure that every slide has something on it that will cue you if your mind goes totally blank in the middle of the talk. You should be able to pop up a random slide, and seamlessly pick up the talk from that point. This is as much a matter of rehearsal as talk design, but you can save yourself a lot of anxiety by making sure that every slide has a clear role in the talk.
11) Structure Matters. Make sure that there is a clear and logical flow to the talk, with one topic leading nicely into the next. The goal is to have the audience leave knowing something about what you did, and our brains are trained to grab hold of narrative. Use that to your advantage, and make your talk into a story-- people will stick with you much better.
You can sort of approximate this by using regular outline slides, but I personally think these are usually a waste of time. If you're stopping every few slides to remind people of where you are and where you've been, you're either confusing your audience, or insulting them.
I tend to use a "take-home messgae" slide at the very beginning that gives the key idea ("We can use laser cooling techniques to measure krypton contamination in ultra-pure neon at the 10-14 level in only a few hours."), to establish the goal of the talk, and then try to design the talk so that the direction is always obvious. I do occasionally remind people of the main point along the way, but I don't make specific outline slides.
That's probably enough rules to make a point of some sort. I'll try to pull together a few files showing the slides I use for various talks, if I can beat them down to a reasonable size for uploading. Otherwise, comments and additions are welcome.